Emission-factor uncertainties in maritime transport in the Strait of Gibraltar, Spain

Abstract. A reliable and up-to-date maritime emission inventory is essential for atmospheric scientists quantifying the impact of shipping. The objective of this study is to estimate the atmospheric emissions of SO₂, NO_x, CO₂ and PM₁₀ by international merchant shipping in 2007 in the Strait of Gibraltar, Spain, including the Algeciras Bay by two methods.

Two methods (both bottom-up) have been used in this study:

1. Establishing engine power-based emission factors (g kWh⁻¹, EPA) or the mass of pollutant per work performed by the engine for each of the relevant components of the exhaust gas from diesel engines and power for each ship.

2. Establishing fuel-based emission factors (kg emitted/t of fuel) or mass of pollutant per mass of combusted fuel for each of the relevant components of the exhaust gas and a fuel-consumption inventory (IMO).

In both methods, the means to estimate engine power and fuel-consumption inventories are the same. The exhaust from boilers and incinerators is regarded as a small contributor and excluded. In total, an estimated average of 1 389 111.05 t of CO₂, 23 083.09 t of SO₂, 32 005.63 t of NO_x and 2972 t of PM₁₀ were emitted from January 2007 until December 2007 by international and domestic shipping. The estimated total fuel consumption amounts to 437 405.84 t. The major differences between the estimates generated by the two methods are for NO_x (16% in certain cases) and CO (up to 23%).

A total difference for all compounds of 3038 t (approximately 2%) has been found between the two methods but it is not areasonable estimate of uncertainty.

Therefore, the results for both methods may be considered acceptable because the actual uncontrolled deviations appear in the changes in emission factors that occur for a given engine with age. These deviations are often difficult to quantify and depend on individual shipboard service and maintenance routines. Emission factors for CO and NO_x are not constant and depend on engine condition. For example, tests conducted by the authors of this paper demonstrate that when an engine operates under normal in-service conditions, the emissions are within limits. However, with a small fault in injection timing, the NO_x emission exceeds the limits (30% higher value in some cases). A fault in the maintenance of the injection nozzles increases the CO emission (15% higher value in some cases).